2.3.1 Degeneration of Cartilage

Articular cartilage is the most critical component of synovial joints, consisting of cells and extracellular matrix (ECM) components [37]. The only cell type present in articular cartilage is the chondrocyte, which is embedded within the ECM, maintaining a low metabolic state. Chondrocytes possess primary cilia and other mechanosensitive receptors on their surface, enabling them to sense and adapt to physical forces, thereby strictly regulating the biochemical composition and structural organisation of articular cartilage [38, 39]. The ECM is composed of water (> 70%) and organic extracellular matrix. The organic extracellular matrix forms a stable network structure primarily composed of type II collagen and several other collagens and non-collagenous proteins, ensuring the tensile strength of cartilage [37, 40]. Various charged proteoglycans embedded in this network structure attract water into the cartilage through their hydrophilic side chains, providing compressive resistance [41].

Articular cartilage is susceptible to external damage, long-term mechanical stimulation, aging and inflammatory responses, all of which contribute to the development and progression of OA [42]. The degenerative changes in articular cartilage and abnormalities in chondrocytes are the core pathological features of OA, primarily characterised by cartilage matrix degradation, chondrocyte apoptosis and cartilage surface damage. In the early stages of OA, chondrocytes exhibit increased synthetic activity, compensatorily secreting type II collagen and proteoglycans (PGs) to repair the damaged matrix [43]. However, as the disease progresses, the activity of MMPs increases, such as MMP-1, MMP-3 and MMP-13, leading to the degradation of type II collagen and the destruction of the pericellular matrix of chondrocytes, resulting in functional inhibition [44]. Simultaneously, a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS)-4/5 promotes cartilage degradation by cleaving aggrecan core proteins and other matrix metalloproteinases [45]. The cartilage loses its hydration capacity, becoming extremely fragile. The depletion of proteoglycans and erosion of the collagen network mark irreversible pathological progression. In OA, the inflammatory response of chondrocytes is significantly enhanced, characterised by increased production of cytokines, chemokines and other pro-inflammatory substances, including interleukin-1β (IL-1β), interleukin-6 (IL-6) and tumour necrosis factor (TNF-α) [46]. Additionally, the autophagy mechanism of chondrocytes declines, leading to increased cell death [47]. As OA progresses, many chondrocytes show increased expression of genes and production of proteins associated with hypertrophy. Angiogenic factors such as vascular endothelial growth factor are also induced, promoting vascular invasion and the expansion of calcified cartilage [48-50]. Changes in cartilage composition lead to significant alterations in its material properties, increasing susceptibility to physical damage and further exacerbating chondrocyte dysfunction and cartilage surface damage [37]. In terms of gross pathological changes, early stage OA patients exhibit surface fibrillation of articular cartilage, which becomes yellowish and loses its lustre. This is followed by localised softening and loss of elasticity [51]. During weight-bearing activities, the cartilage surface becomes rough and undergoes fragmented peeling. Deep fissures in the articular cartilage and the formation of intra-articular loose bodies eventually lead to cartilage shedding, exposure of the underlying calcified cartilage and subchondral bone [52].

2.3.2 Subchondral Bone Remodelling

The anatomical composition of subchondral bone includes the subchondral bone plate (cortical plate) and subchondral trabecular bone, forming the tidemark as the mechanical interface with cartilage. Subchondral bone undergoes complex, dynamic structural remodelling and molecular regulation imbalances in OA, with pathological changes involving anatomical abnormalities, cellular dynamic imbalances and mechanical conduction disturbances [53, 54]. The subchondral bone cortical plate thickens and bone mineralisation status changes. Weinans et al. noted that although subchondral bone volume increases in OA, bone hardness decreases due to reduced bone mineral density [55]. Trabecular bone undergoes remodelling, with decreased trabecular thickness, increased spacing and microcrack formation. Gallant et al. indicated that these trabecular changes are directly related to increased mechanical load [56]. Additionally, characteristic anatomical pathological markers form in subchondral bone, including subchondral cysts resulting from osteoclast-mediated bone resorption and fluid accumulation, as well as marginal osteophytes formed through endochondral ossification [57]. Pelletier et al.'s MRI imaging studies suggested that increased subchondral cyst size correlates with cartilage volume loss in the same region, highlighting the importance of subchondral bone lesions in OA pathophysiology [58]. Donald et al. conducted MRI imaging studies showing that bone marrow lesions and micro-fractures coexist in subchondral bone, indicating an imbalance in the dynamic equilibrium between local repair and destruction [59]. Abnormal bone remodelling in subchondral bone is driven by functional imbalances among osteoblasts (OBs), osteoclasts (OCs) and endothelial cells (ECs) [54]. Su et al. identified three OBs subtypes involved in abnormal subchondral bone remodelling through single-cell sequencing: endothelial osteoblasts (promoting angiogenesis via VEGF signalling), stromal osteoblasts (leading collagen fibre assembly) and mineralising osteoblasts (mediating abnormal mineralisation as terminally differentiated cells, leading to increased bone fragility) [60]. The vascular-osteogenic coupling mechanism also promotes abnormal subchondral bone remodelling [61]. Shiwu et al. found that pre-endothelial cells form functional coupling with OBs through the TGFβ/VEGF/NOTCH pathway [54]. The differentiated H-type ECs not only promote pathological vascular invasion but also activate osteoclasts by releasing receptor activator of nuclear factor-κ B ligand (RANKL) [62], further recruiting bone progenitor cells through the Smad2/3 pathway to accelerate abnormal bone turnover, forming a ‘vascular-osteogenic-osteoclastic’ vicious cycle. This cellular network dysregulation leads to abnormal bone turnover rates, manifested as osteoid island deposition and osteophyte formation [63]. In the early stage, abnormal vascularisation dominates. Christelle et al. noted that increased mechanical load triggers OBs to secrete inflammatory mediators such as IL-6 and MMP-13 [64], while H-type vessels penetrate the tidemark, with MRI revealing bone marrow lesions and micro-fractures [59]. In the middle stage, characteristic subchondral cysts form. Michel et al. indicated that their formation mechanism is related to osteoclast-mediated bone resorption and fluid accumulation, with reduced cortical plate hardness exacerbating cartilage deformation [65]. In the late stage, marginal osteophyte proliferation occurs through endochondral ossification. Although this provides compensatory joint stability, it alters joint mechanical distribution [66, 67]. Notably, Felson et al. pointed out that osteophyte formation is closely related to pain, possibly due to the rich innervation of subchondral bone [67]. Studies have shown that both bone sclerosis and softening accelerate cartilage degeneration by altering stress distribution [68, 69]. The molecular basis of bone–cartilage interaction provides new therapeutic targets for OA, such as anti-angiogenic therapy targeting H-type vessels, interventions in Pre-ECs differentiation to block pathological bone remodelling or epigenetic strategies to regulate OBs differentiation. However, current research still faces technical limitations, such as the loss of osteoclast phenotypes during single-cell isolation, necessitating methodological breakthroughs [60].

2.3.3 Synovial Inflammatory Hyperplasia

The synovial membrane is located on the inner layer of the joint capsule, covering the ligaments and tendons within the joint. It consists of an outer layer of synovial epithelial cells and an inner layer of synovial fibroblasts. Physiologically, the cellular components of the synovial membrane secrete synovial fluid, which contains a high concentration of hyaluronic acid, playing a crucial role in lubrication during joint movement. During the pathological progression of OA, significant structural and functional alterations occur in the synovium, which are closely associated with symptoms such as pain and synovitis. Synovial inflammation represents a critical aspect of OA progression, characterised by pathological features including fibrosis, macrophage infiltration and synovial epithelial hyperplasia [70]. Synovial inflammation exhibits distinct pathological characteristics at different stages of OA. In early-stage OA, synovial inflammation typically manifests as mild hyperplasia and macrophage infiltration. As the disease progresses, synovial inflammation intensifies, presenting with pronounced fibrosis and synovial hyperplasia [71]. This inflammatory response is not confined to the synovial layer but may also affect sub-synovial tissues, leading to the accumulation of local immune cells and the release of pro-inflammatory factors [72]. Synovial fibrosis is one of the key pathological features of synovial inflammation in OA. Research indicates that synovial fibroblasts (SFs) play a significant role in OA by secreting collagen and MMPs, thereby promoting cartilage degradation and synovial fibrosis [73]. Additionally, macrophages and T cells within the synovium contribute to the inflammatory response by releasing pro-inflammatory factors such as IL-6 and TNF-α, further exacerbating synovial pathological changes [74] (Figure 2).

4.1.1 Pluripotent Stem Cells

PSCs encompassing ESCs and iPSCs [124], can precisely recapitulate bone and cartilage development from the ectoderm stage onward through directed differentiation strategies. Among these, ESCs—established as the earliest pluripotent cell line—display strong developmental fidelity and notable in vitro stability and scalability, making them an ideal tool for organoid construction [125, 126]. ESCs are capable of differentiating into multiple types of organoids, including those of the brain [127], liver [128], heart [129] and kidney [121]. However, obtaining ESCs is relatively challenging because it requires the use of early-stage embryos or trophoblastic cells. The involvement of early embryos raises significant ethical concerns, which limit the application of ESCs in organoid construction [130]. iPSCs represent a groundbreaking advancement in tissue engineering, as they enable the reprogramming of adult cells into a pluripotent state through transcription factor induction, granting them the ability to differentiate into various cell types [131]. iPSCs can differentiate into organ-specific organoids, mimicking the complex structure and function of native organs. They also offer advantages such as a broad donor availability, lack of ethical concerns and stable genetic integrity. The advent of iPSCs, achieved through somatic cell reprogramming via defined transcription factors, has revolutionised the field by maintaining equivalent pluripotency while circumventing the embryo-derived ethical controversies inherent to ESCs utilisation. However, it is important to acknowledge that while iPSCs eliminate the need for embryonic materials, ethical considerations persist regarding their biospecimen sourcing protocols and translational applications. These include ensuring informed consent for somatic cell donations, addressing potential commercialisation challenges of personalised cell lines and establishing regulatory frameworks for clinical-grade iPSCs derivation. This evolving ethical landscape necessitates continued discussions to balance scientific progress with responsible innovation as iPSC-based organoid technologies advance toward therapeutic implementations [132, 133].

iPSC-based organoid research has made significant strides in disease modelling, drug development and regenerative medicine. Organoids derived from iPSCs have been successfully developed for vascular [134], pulmonary [135], cardiac [129] and intestinal tissues [136], among others. iPSCs also serve as an optimal cellular source for arthrosis organoid construction. Under specific growth factors and 3D culture systems, iPSCs can efficiently differentiate into chondrocytes, producing an extracellular matrix rich in glycosaminoglycans (GAGs) and type II collagen. Additionally, iPSCs can be directed into osteogenic differentiation through mesodermal induction protocols, upregulating bone marker genes such as COL1A1 and BGLAP and forming mineralised matrices [137]. Shannon et al. successfully developed bone and cartilage organoids from mouse-derived iPSCs, replicating endochondral ossification and osteogenesis [138]. Gabriella et al. demonstrated that iPSC-derived chondrocytes could form cartilage-like nodules under scaffold and scaffold-free conditions, integrating with host cartilage to facilitate defect repair, thereby validating the self-organising properties of cartilage organoid formation [139]. Khan et al. induced iPSCs to differentiate into mesenchymal, endothelial and haematopoietic lineage cells, successfully replicating the complex microenvironment of human bone marrow [140]. OA organoids consist of a cartilage–subchondral bone–synovium composite structure, requiring differentiation-specific regulatory strategies for seed cells to guide stem cell-directed differentiation into cartilage, subchondral bone and synovial layers. Cartilage layer induction can be achieved using iPSCs combined with TGF-β3 and BMP-2 to establish a cartilage-specific differentiation system, activating the SOX9/COL2A1 pathway for chondrogenic regulation [141, 142]. Subchondral bone layer induction can be performed by dual-signal activation of the BMP-2/BMP-7 osteogenic differentiation pathway, coupled with mechanical loading to promote mineralisation [143, 144].

4.1.2 Mesenchymal Stem Cells

MSCs, with their multipotent differentiation potential and ease of isolation, are highly valuable for constructing bone and cartilage organoids [145]. MSCs are widely found in bone marrow, adipose tissue, umbilical cord, placenta and dental pulp. Under specific induction conditions (e.g., TGF-β, BMP-6), they can differentiate into osteoblasts or chondrocytes and secrete key matrix components such as osteocalcin and type II collagen [146-148]. In bone organoid construction, BMSCs are often combined with bioactive materials (e.g., bone matrix-inspired bioinks, calcium phosphate ceramic scaffolds) to enhance bone formation and vascularisation through self-mineralisation or co-delivery of endothelial cells/growth factors [149]. Su et al. successfully constructed self-mineralising bone organoids using BMSCs and bone matrix-induced bioinks, which exhibited natural bone-like structures after transplantation in nude mice [99]. Martin et al. recreated the bone marrow microenvironment using BMSCs and human haematopoietic stem cells (HSCs) [150]. Additionally, periosteum-derived stem cells are preferred for bone repair due to their sensitivity to mechanical stimuli, while adipose-derived stem cells are gaining attention for their abundance, low immunogenicity and stable proliferation [151]. Cartilage organoid construction relies on high-density culture and microenvironment simulation techniques. For example, GelMA microspheres or 3D scaffolds combined with TGF-β3 induction can promote MSC differentiation into functional chondrocytes and the formation of layered cartilage structures. Although BMSCs remain the primary choice, adipose-derived MSCs show comparable chondrogenic differentiation efficiency and are easier to obtain [90]. Notably, the immunomodulatory function of MSCs can improve the local microenvironment by inhibiting inflammatory factors (e.g., IL-1β in osteoarthritis), indirectly promoting organoid maturation [148, 152]. However, differentiation efficiency and stability vary among MSC sources. For example, dental pulp stem cells excel in vascularised dental pulp organoid construction but require optimised pre-differentiation strategies to avoid proliferation inhibition [149, 153].

4.1.3 Haematopoietic Stem Cells and Their Progeny

HSCs are a unique population within the bone marrow, possessing both self-renewal and multipotent differentiation capacities that enable the continual regeneration of all lineages in the adult blood and immune systems [154, 155]. They retain this distinctive self-renewal ability throughout the lifespan, making them the key cell type responsible for maintaining haematopoietic homeostasis [156]. Among the progeny of HSCs are various lineage progenitor cells—including myeloid progenitor cells (MkPs), lymphoid progenitor cells (EpPs) and endothelial progenitor cells (EPCs)—which under specific conditions can further differentiate into mature blood cells such as erythrocytes, leukocytes, platelets and vascular endothelial cells [157, 158]. By culturing HSCs and their progeny in vitro, researchers can construct blood organoids that recapitulate human haematopoiesis for disease modelling and drug screening. Additionally, these cells can be used to build immune organoids, thereby facilitating the study of immune system development and function [159]. EPCs are a pluripotent cell population capable of differentiating into mature ECs, playing a crucial role in vascular regeneration and repair. EPCs can be classified into early EPCs and late EPCs: Early EPCs exhibit higher proliferative capacity, whereas late EPCs display more mature endothelial characteristics [160]. Biphasic differentiation of EPCs refers to their ability to differentiate into synovial fibroblasts or endothelial cells, regulated by multiple signalling pathways [161]. Differentiation into synovial fibroblasts is associated with Notch signalling activation, which has been shown to promote EPC-to-synovial cell differentiation in both in vivo and in vitro studies [162]. In contrast, differentiation into endothelial cells is primarily driven by VEGF and SDF-1, which enhance EPCs migration and proliferation, leading to the formation of a mature endothelial network [161]. By precisely controlling the synovial–vascular lineage biphasic differentiation of EPCs, it is possible to generate mature synovial fibroblasts and endothelial networks, facilitating the construction of the synovial layer in OA organoids. Synovial layer construction can be facilitated through directed induction of EPCs by TGF-β1 and PDGF-BB, generating synovial fibroblasts and vascular endothelial cells, while incorporating macrophages to simulate the inflammatory microenvironment [163].

4.1.4 Functional Cells

In the preparation of OA organoids, functional cell populations beyond stem cells are also pivotal contributors by imparting distinct pathological environments. Derived from MSCs, osteoblasts synthesise and mineralise the bone matrix, secreting key proteins such as collagen and osteocalcin [164]. In organoid construction, osteoblasts can be obtained by inducing the MC3T3-E1 cell line differentiation and participate in trabecular bone formation [165]. Originating from HSCs (e.g., monocyte/macrophage lineage), osteoclasts differentiate via the RANKL-RANK signalling pathway [164]. Osteoblast-secreted RANKL and macrophage colony-stimulating factor are critical for osteoclast differentiation [166]. The interplay between osteoblasts and osteoclasts is essential for bone repair, regeneration and the study of osteoporosis and osteoarthritis [167]. Metcalfe et al. constructed micrometre-scale bone organoids where osteoclast overactivation simulated pathological bone loss (e.g., osteoporosis) and co-culture systems with osteoblasts modelled bone remodelling imbalances [96]. Comprising 90%–95% of adult bone cells, osteocytes are embedded in the mineralised matrix and regulate osteoblast/osteoclast activity through dendritic networks, acting as mechanical stress sensors [166, 168]. Incorporating osteocytes enhances organoid physiological relevance by simulating bone microenvironment signalling (e.g., RANKL, OPG) [169]. Chondrocytes maintain cartilage structure by secreting type II collagen and aggrecan and differentiate into osteoblasts during endochondral ossification [165, 170]. Co-culturing chondrocytes with endothelial cells can simulate the articular cartilage–bone interface, useful for studying cartilage degeneration and ectopic ossification in osteoarthritis [171]. Simulating pain perception additionally relies on the incorporation of neurons within the OA organoids [172]. Notably, macrophages and other immune cells function as key contributors within OA organoids, facilitating essential immune responses and shaping disease-related processes. M1 macrophages are the primary pro-inflammatory cells in synovitis, activated through multiple signalling pathways (e.g., NF-κB, MAPK) and secreting large amounts of pro-inflammatory cytokines, including IL-1β, TNF-α, IL-6, IL-12, CCL2 and CCL5 [173]. These cytokines not only sustain the inflammatory response but also exacerbate inflammation through interactions with T cells, neutrophils and monocytes [174]. Additionally, M1 macrophages release reactive oxygen species and nitric oxide, directly damaging joint tissues. They also interact with synovial fibroblasts, further promoting inflammatory cytokine secretion and overexpression of inflammatory mediators [175]. In synovial inflammation, the polarisation of M1 macrophages and the abnormal activation of synovial fibroblasts play key roles in cytokine cascade release, MMP overexpression and inflammatory mediator upregulation. Incorporating macrophages and synovial cells in OA organoids enhances the simulation of the OA inflammatory microenvironment.

4.2.1 Matrix Materials for Cartilage Tissue Engineering

The matrix materials required for cartilage tissue engineering include various natural and synthetic hydrogels such as collagen, gelatin and hyaluronic acid (HA), as well as polyethylene glycol (PEG), poly(lactic-co-glycolic acid) (PLGA) and bioactive molecules. Through different processing methods and combinations, these materials can form highly biocompatible and bioactive scaffolds, providing a diverse selection for cartilage tissue engineering. Gelatin, a partial hydrolysis product of collagen, retains the arginine-glycine-aspartic acid (RGD) sequence. It provides dynamic mechanical stimulation in hydrogel form, promoting cell migration and matrix deposition and exhibits excellent biocompatibility and ECM similarity, making it widely used in cartilage organoids construction [179]. Yang et al. developed gelatin-based micro-cryogels that achieved synchronous regeneration of cartilage and bone tissues in organoids, with degradation rates matching tissue regeneration speeds [180]. Additionally, hydrogels are widely used in cartilage tissue engineering [181, 182]. Su's team highlighted that silk fibroin-based hydrogels not only possess an ECM-like structure but also exhibit unique mechanical properties and excellent biocompatibility, making them an ideal biomaterial for cartilage tissue construction [183]. Microspheres prepared from hydrogels offer tunable physical, chemical and biological properties, effectively mimicking the 3D ECM-like scaffold required for cartilage growth. In particular, Su et al. developed RGD-silk fibroin-DNA hydrogel microspheres (RSD-MS), which demonstrated favourable material properties and significantly enhanced cartilage regeneration in in vivo studies [92]. Hydrogels can also serve as carriers for multiple bioactive factors and exosomes, enabling them to simulate various physiological and pathological environments. Zhang et al. utilised 3D bioprinting hydrogel combined with exosomes loaded with chondrogenic stimuli agents, which exhibited notable immunomodulatory effects. This faster and safer biomaterial promotes cartilage proliferation and repair, making it highly applicable in the treatment of osteoarthritis and other cartilage-related diseases [184]. PLGA is a biodegradable copolymer composed of lactic acid and glycolic acid. With excellent biocompatibility and mechanical strength, PLGA provides a robust structural environment conducive to organoid development. Its degradation rate can be fine-tuned by adjusting the lactic acid-to-glycolic acid ratio, ensuring synchronisation with cell growth and differentiation. Chen et al. developed PLGA matrix loaded with a cell-free adipose extract, which effectively facilitated the formation of cartilage organoids and promoted chondrocyte differentiation [185].

4.2.2 Matrix Materials for Subchondral Bone Tissue Engineering

Appropriate matrix materials play a crucial role in actively inducing subchondral bone mineralisation and osteogenesis, providing essential support for bone tissue formation and function. These materials are indispensable for developing functional subchondral bone tissue. A wide variety of matrix materials are used for subchondral bone construction, which are similar to the matrix materials for cartilage tissue engineering, including collagen, Matrigel, hydrogels, alginate, hydroxyapatite, chitosan, HA, bioactive glass, polylactic acid and microporous materials. Collagen is the primary organic component of the bone ECM, with type I collagen comprising approximately 90% of the total bone matrix. It is frequently used as a scaffold base to mimic the 3D bone microenvironment and promote MSCs differentiation into osteoblasts. Matrigel, derived from the basement membrane of Engelbreth-Holm-Swarm mouse sarcoma, primarily consists of laminin, collagen IV, heparan sulphate proteoglycan (perlecan) and various growth factors such as fibroblast growth factor, epidermal growth factor and transforming growth factor-β [186]. In a study by Enyi et al., Matrigel was used to support BMP9-induced osteogenic differentiation and bone matrix mineralisation, demonstrating its significance in alveolar bone cell differentiation and bone tissue engineering [187]. These components form a three-dimensional network structure similar to natural ECM, providing an excellent environment for cell attachment and growth, and supporting 3D culture and osteogenic/chondrogenic differentiation of stem cells [188]. Hydrogels, due to their excellent biocompatibility and tunable mechanical properties, have emerged as key materials in bone tissue engineering, serving as scaffolds to support cell growth and differentiation. GelMA hydrogels, known for their adjustability and biocompatibility, have been extensively studied for musculoskeletal tissue regeneration [189]. For instance, self-assembling peptide hydrogels exhibit exceptional performance in bone regeneration, facilitating cell adhesion, migration, proliferation and differentiation [190]. Su et al. developed an ECM-DNA-CPO-based engineered biomimetic hydrogel with a dynamic network to simulate the bone microenvironment, facilitating the construction of vascularised and mineralised bone organoids [191]. The 3D printing technology has played a significant role in the fabrication and application of hydrogel-based materials. By leveraging 3D printing, researchers can integrate extracellular matrix components within hydrogels, creating adequate spatial structures for osteogenesis and chondrogenesis, thereby promoting tissue repair [192]. Additionally, DNA hydrogels, recognised for their programmability, tunability and mechanical properties, are considered a promising innovation in bone tissue engineering. Su's team highlighted the potential of DNA-functionalised bio-ink combined with 3D bioprinting to construct bone tissue organoids, showcasing its prospects in bone regeneration and osteoarthritis treatment [193]. PEG is a highly biocompatible, water-soluble polymer characterised by exceptional hydrophilicity and biological inertness, making it widely applicable in tissue engineering and organoid construction [194]. Chemical modifications, such as the introduction of reactive functional groups, enhance PEG's interactions with cells and allow for mechanical property adjustments, thereby improving its practical utility in tissue engineering. Ehrbar et al. developed a hybrid hydrogel by combining PEG with HA and utilising a transglutaminase crosslinking mechanism. This significantly improved the physical and biochemical properties of the hydrogel, providing an optimal microenvironment for the development of humanised bone marrow organoids [150].

4.2.3 Matrix Materials for Synovial Tissue Engineering

The key to constructing synovial tissue lies in accurately mimicking its natural microenvironment, where the selection of matrix materials—particularly hydrogels—directly influences cellular behaviours such as proliferation, differentiation and functional expression. The choice of synovial tissue matrix materials must meet biomechanical requirements, enabling adaptation to mechanical compression, loading and movement-related forces acting on synovial tissue. By designing scaffold materials with specific mechanical properties, researchers can promote synovial cell growth and differentiation while effectively simulating the natural mechanical environment of synovial tissue [195]. Natural hydrogel materials include collagen, fibrin and hyaluronic acid, while synthetic materials such as polylactic acid and PLGA, along with composite materials like chitosan/hydroxyapatite, are also commonly used. As the core matrix material for synovial tissue, hydrogels must balance biomimicry, tunability and functionality. Guo et al. utilised synovial MSCs as seed cells and employed a chitosan hydrogel/3D-printed poly(ε-caprolactone) hybrid scaffold, combined with tetrahedral framework nucleic acid recruitment, to successfully regenerate both cartilage and synovial tissue [196]. Dai et al. developed a senescence-targeted miR-24 μS/SMSC organoid hydrogel for synovial organoid construction. This material demonstrated excellent chondrogenic potential in vitro, while animal experiments showed superior cartilage repair at 24 weeks, better joint function maintenance and reduced intra-articular inflammatory response following rat joint transplantation [197]. Peter et al. successfully constructed a synovial organoid-on-a-chip using a composite matrix composed of Matrigel, PEG–dextran and fibrin hydrogels. This platform enables monitoring of tissue-level remodelling under arthritic pathological conditions [198]. Future research will focus on dynamically responsive materials, multi-scale structural design and clinical translation, providing a more reliable material foundation for precise synovial tissue construction.

4.3.1 Inflammatory Type of Osteoarthritis Organoids

The inflammatory OA organoid model is primarily driven by immune system dysregulation and chronic inflammation, characterised by immune cell infiltration and a pro-inflammatory cytokine cascade. The core mechanisms involve T-cell, B-cell and macrophage-mediated synovial inflammation, leading to abnormal secretion of pro-inflammatory factors such as IL-1β and TNF-α, which in turn trigger cartilage matrix degradation (upregulation of MMP-13 and COX-2), joint effusion and accelerated chondrocyte apoptosis, ultimately resulting in cartilage destruction and joint stiffness [201]. Using iPSCs co-cultured with a matrix gel embedded with pro-inflammatory factors makes it possible to emulate an inflammatory microenvironment. Currently, Matrigel loaded with inflammatory factors—such as programmable DNA hydrogels, nanocomposite hydrogels and smart responsive hydrogels—has attracted significant research attention [89]. Notably, programmable DNA hydrogels exhibit unique programmability compared with other natural and synthetic polymer hydrogels, allowing precise structural customisation and tunable properties [202]. Targeted DNA segments can be generated via hybridisation chain reactions, PCR amplification or other methods and subsequently assembled stepwise into an ordered DNA-based matrix under electrostatic interactions. Through careful sequence design, researchers can tailor the matrix's physicochemical properties—such as rigidity, distribution of active sites, topological texture and creep behaviour—thus achieving fine control over the hydrogel's characteristics [203]. By combining programmable DNA hydrogels with other matrix materials, it is possible to effectively mimic cartilage, subchondral bone and synovial matrices to regulate the inflammatory differentiation of stem cells. Specifically, inducing inflammation in stages typically entails a multi-step cascade. First, a low concentration of pro-inflammatory factors is applied to initiate a baseline inflammatory response. Subsequently, additional factors are introduced stepwise to replicate progressively more complex inflammatory environments. Treating cells with low-concentration IL-1β for 24 h induces a baseline inflammatory response. Subsequently adding TNF-α and IL-6 simulates a progressively complex inflammatory microenvironment. Keishi et al. developed an inflammatory microenvironment model for RA fibroblast-like synoviocytes by using a stepwise cytokine induction approach, where cells were sequentially stimulated with TNF-α, IL-1β and IL-6, effectively recapitulating the inflammatory cascade of RA in vitro [204].

4.3.2 Mechanical Stress Type of Osteoarthritis Organoids

Mechanical stress-induced OA is primarily caused by excessive or abnormal mechanical loading. Prolonged mechanical overload disrupts the biomechanical balance at the cartilage–bone interface through abnormal stress transmission, potentially leading to microcracks or fractures on the articular cartilage surface, which compromises cartilage structure and load-bearing capacity, ultimately accelerating cartilage degeneration [205]. Additionally, mechanical pressure contributes to the degradation of cartilage matrix components, such as collagen fibre rupture and proteoglycan loss, while simultaneously inhibiting the repair process, resulting in metabolic imbalance and ultimately leading to joint pain, swelling and functional impairment [206]. The construction of mechanical stress-induced OA organoids relies on dynamic culture systems, utilising multi-axial bioreactors to apply cyclic mechanical stress. Load parameters are set, allowing precise control over stress magnitude, frequency and duration. This approach enables researchers to investigate the effects of varying degrees of mechanical stress on osteocytes, chondrocytes and bone organoids, ultimately defining the optimal mechanical environment that induces OA. By systematically adjusting these parameters, researchers can better simulate the physiological and pathological biomechanical environment of joints. A combination of dynamic compression and dynamic shear loading is employed to better replicate the physiological mechanical state of cartilage tissue in vivo. Lisbet et al. conducted studies on high mechanical strain in 3D matrices, aiming to elucidate the biomechanical mechanisms underlying OA progression. Their research analysed gene expression changes in chondrocytes subjected to mechanical dynamic strain stimulation, revealing how mechanical stress influences cellular behaviour [172]. Additionally, electro-mechanical coupling should be enhanced during culture to augment mechanical stress transmission within OA organoids. This involves the application of electrical stimulation or magnetic field reinforcement, such as low-intensity pulsed ultrasound or static magnetic fields, to activate Piezo1 ion channels and promote chondrocyte mechanotransduction. These strategies maximise the effects of mechanical stress on OA organoids. Geng et al. demonstrated that LIPUS generates cyclic mechanical waves that improve ischaemia, suppress inflammation and promote cartilage repair. Moreover, LIPUS significantly enhances chondrocyte proliferation and extracellular matrix synthesis while reducing osteoclast activity, further contributing to OA prevention and treatment [207].

4.3.3 Genetics Type of Osteoarthritis Organoids

Genetic predisposition OA involves multifactorial interactions, with gene mutations or polymorphisms playing a pivotal role, alongside factors affecting chondrocyte differentiation, inflammatory responses and extracellular matrix degradation. Specifically, polymorphisms in genes such as ESR1 and GDF5 are closely associated with an increased risk of knee OA, while variations in cytokine-related genes (e.g., IL-1β and TGF-β1) further contribute to OA susceptibility and severity [208]. Additionally, dysregulated immune cell function, cytokine imbalances (e.g., abnormal secretion of TNF-α and IL-17) and autoantibody production collectively drive joint inflammation and cartilage destruction [209]. Environmental factors can act as triggers in genetically predisposed individuals, accelerating cartilage degeneration. The core of genetic predisposition organoid models lies in gene editing strategies. The process involves selecting OA-related target genes, such as RUNX1, HIF-2α, MAPK12, FOS, MMP-13 and Sox9, and designing specific single guide RNAs (sgRNAs) based on target gene sequences. The Cas9 protein and sgRNAs sequences are then inserted into expression plasmid vectors, with adeno-associated viruses (AAV) or lentiviruses serving as delivery systems to introduce the CRISPR/Cas9 system into target stem cells. This enables the knockout of cartilage-protective genes or overexpression of pathogenic genes. Chen et al. utilised CRISPR/Cas9-mediated gene knockout and downregulation of MMP-13, IL-1β and NGF, significantly mitigating OA-associated inflammation and cartilage degradation [210]. Kiran et al. replaced TALEN with CRISPR technology, enhancing the efficiency of genome editing in hPSCs and enabling the development of genotype-specific organoids [211].

4.4.1 3D Bioprinting Technology

3D bioprinting technology represents a revolutionary advancement in tissue engineering, enabling the precise spatial control of cells, biomaterials and bioactive factors to construct complex 3D structures [212]. Bioprinting allows for the fabrication of high-fidelity arthrosis organoids, with key advantages including high precision and reproducibility in tissue construction [213]. Using high-precision DLP bioprinting technology, it is possible to layer-by-layer print cartilage–subchondral bone–synovium composite structures with exceptionally high spatial positioning accuracy of cells [214]. The stability of organoid structure and function depends on the precise control of multiple factors, including biomaterial preparation, printing speed, temperature and layer thickness [215]. Lode et al. successfully achieved full-thickness osteochondral tissue regeneration by employing bioprinting technology with bioinks containing primary bone–cartilage substitute cells, in combination with multi-layered mineralisation constructs [216]. Additionally, 3D bioprinting technology allows for the deposition of bioinks containing BMSCs, ECM and growth factors, integrating multiple biomaterials (e.g., GelMA, AlgMA and hydroxyapatite) to further optimise the structure and function of organoids, achieving mechanical properties highly similar to natural bone tissue. Ouyang et al. developed bone organoids with reparative properties by encapsulating BMSCs in GelMA microspheres using 3D bioprinting, followed by stepwise differentiation induction [87]. Similarly, Su et al. applied 3D bioprinting technology to fabricate self-mineralising bone organoids using bioinks containing a diverse range of biomaterials [99]. The introduction of nanoparticles provides additional advancements in bioprinting technology, enhancing its capability to construct complex tissue structures. Specifically, nanoparticles improve the precision and functionality of bioprinted tissues, facilitating the accurate replication of the intricate architecture of bone and cartilage [216].

4.4.2 Microfluidic Technology

The application of microfluidic and microporous array chip technology in organoid manufacturing represents a significant advancement in biomedical engineering. By leveraging precisely designed microfluidic chips and modular construction, this technology enables dynamic control of the cell culture environment, effectively mimicking the pathophysiological characteristics of OA while enhancing organoid culture consistency and efficiency. Abbasalizadeh et al. utilised droplet microfluidic technology to continuously generate cell-laden microcapsules, promoting cell proliferation, self-contraction and self-organisation into organoids, ultimately leading to large-scale development of fully structured liver organoids [217]. Microfluidic and microporous array chips function by creating micro-scale pores or channels within the chip, allowing for precise simulation of the human tissue microenvironment. By controlling the size and distribution of these pores, researchers can finely regulate intercellular interactions and signal transduction [218], thereby effectively replicating the inflammatory cell interactions and cytokine cascades observed in the pathological microenvironment of OA. Moreover, microfluidic systems can simulate biological processes across multiple organ systems, including vascular, neural, respiratory and digestive systems. For example, microfluidic models have successfully reconstructed the alveolar-capillary interface, replicating its role in respiration and inflammation [219]. This technology provides crucial support for the construction of cartilage–subchondral bone–synovium interfaces in OA organoid models. Additionally, microfluidic technology facilitates continuous and dynamic organoid culture platforms, enabling researchers to real-time monitor organoid behaviour and drug responses, thereby optimising drug screening and disease modelling. Khademhosseini et al. proposed a fully integrated modular system incorporating physical, biochemical and optical sensing platforms, which operates organ-on-a-chip units in a continuous, dynamic and automated manner [220]. Regarding biomaterial preparation, microfluidic technology, when combined with other advanced fabrication techniques, enables the development of stem cell culture environments with superior ECM biomimetic properties for OA organoid formation. Su et al. employed photo-crosslinking and self-assembly microfluidic integration systems to generate novel RSD-MS, which served as the foundation for constructing primary cartilage organoids [92].

4.4.3 Spheroid Technology

The application of spheroid technology in organoid construction represents a major advancement in tissue engineering. The core principle lies in utilising the 3D structural properties of spheroids to simulate the microenvironment of human tissues, making it a foundational technology for organoid construction. This is particularly valuable for modelling complex cell–cell and cell–matrix interactions, facilitating advancements in disease modelling, drug screening, tissue engineering and regenerative medicine. By suspending single-cell layers or multiple cell types in culture medium, cells spontaneously aggregate into 3D spheroids through physical forces such as gravity, surface tension or magnetism. Sequential application of specific cytokines (e.g., FGF, EGF, WNT-3a, collagen and TGF-β) gradually induces cell differentiation, mimicking the natural developmental progression from mesoderm to mature cartilage, bone and synovium [221, 222]. Martin et al. cultivated spheroids on low-adhesion surfaces and subsequently loaded them into an assembly chamber, where acoustic nodal assembly technology was used to generate hepatic spheroids. These spheroids matured in fibrinogen gel, eventually forming larger organ-like structures [223]. Spheroids recapitulate key aspects of the human tissue microenvironment, including ECM deposition, intercellular interactions and metabolic waste clearance, thereby enhancing the study of cellular behaviour and signal transduction mechanisms [224]. This capability is particularly well-suited for reproducing the pathological microenvironment of OA. Moreover, spheroid technology integrates closely with 3D bioprinting, where high-cell-density bioprinting of spheroids significantly reduces the maturation time required for large tissue/organoid formation [225]. By layered induction of spheroids corresponding to cartilage, subchondral bone and synovium, followed by high-precision DLP bioprinting, it is possible to assemble pathological cartilage–subchondral bone–synovium composite structures. This integration achieves a structurally and functionally unified OA organoid model, effectively replicating OA pathological features in vitro.

5.1.1 Artificial Intelligence

In the ‘human-free factory’ OA pathological organoid intelligent manufacturing strategy, AI serves as the core computational engine. AI refers to computational systems that simulate human cognitive functions, including learning, reasoning, perception and decision-making. Key AI subfields include: Machine learning (ML): Extracts patterns from large-scale datasets for prediction and classification; Deep learning (DL): Uses multi-layer neural networks to model complex relationships, widely applied in image recognition and biomedical data processing; Natural language processing (NLP): Analyses and generates scientific text for literature mining and hypothesis generation. These AI technologies are extensively applied in biomedical research, facilitating data analysis, model development and experimental optimisation [231]. Integrating AI into organoid manufacturing enables AI-assisted construction, analysis and application, significantly enhancing efficiency, precision and reproducibility in organoid production [232]. AI offers a broad range of applications in OA organoid manufacturing, including: Optimisation of culture conditions: AI algorithms analyse large experimental datasets to predict optimal culture parameters, such as temperature, humidity and nutrient concentrations. Additionally, AI can predict cell behaviour and optimise material properties, further enhancing the efficiency and precision of organoid manufacturing. AI-driven optimisation enables rapid screening of existing construction strategies, refining experimental designs to enhance efficiency, accuracy and quality [233]; Image analysis and feature extraction: AI can rapidly extract multi-scale features from organoid microscopic images, improving analysis efficiency. For example, convolutional neural networks analyse single-cell imaging and histological sections, enabling multi-layer, multi-perspective structural insights for precise morphological evaluation [234]; Multi-omics data integration: AI facilitates the fusion of genomics, transcriptomics and proteomics data, supporting streamlined analysis workflows for OA disease modelling [235]; Clinical translation and drug evaluation: In combination with iPSCs, AI enables the generation of patient-specific in vitro disease models, advancing personalised medicine through high-throughput drug screening, efficacy prediction and organoid-based therapeutic assessments [236]. Su et al. introduced the world's first domain-specific AI model for organoids, Organoid-GPT (O-GPT), which explores synergistic effects between intelligent manufacturing and organoid construction. O-GPT provides state-of-the-art domain knowledge, accurate predictions of organoid construction factors and optimised model-data analysis pipelines, offering solutions for multi-scale imaging evaluation and multi-omics data fusion analysis. This framework establishes a virtual research platform for intelligent manufacturing of pathological OA organoids. Additionally, Bai et al. provided a detailed overview of the fundamental concepts and mechanisms by which AI supports organoid construction, summarising its prospective applications in high-throughput screening of organoid fabrication strategies, cost-effective extraction of multi-scale imaging features, simplified multi-omics data analysis and precise preclinical evaluation and application [232]. Specifically, AI can drive precise adjustments to the biochemical and biophysical properties of specific hydrogels during organoid fabrication, greatly improving the final organoid formation [237]. Furthermore, AI can predict and identify stem cell differentiation pathways in organoid cultures. Zhu et al. employed deep learning to extract details from large-scale datasets and developed a deep neural network model for predicting and reliably identifying neural stem cell fate. This model demonstrated high efficiency in recognising differentiated cell types and exhibited remarkable accuracy and robustness across independent test scenarios involving various inducers, including neurotrophic factors, hormones, small-molecule compounds and nanoparticles [238].

To fully implement AI-driven automation, three core software systems must be developed for ‘human-free factory’ OA organoid manufacturing: Automated Culture Optimisation and Real-Time Monitoring System; Multi-Modal Image Database and Image Processing System; Multi-Omics Deep Neural Network System. These three systems function sequentially, interconnecting upstream and downstream processes in a relay-style workflow, collectively powering the ‘human-free factory’ intelligent manufacturing system.

5.1.2 Real-Time Monitoring Feedback System

AI-driven algorithms can analyse complex datasets and predict critical culture variables, including temperature, oxygen concentration, pH levels, biological aging, nutrient composition and growth factors. By leveraging machine learning models, AI determines the optimal conditions for cell growth, differentiation and organoid formation. This approach enables rapid screening of various biomaterials, growth factors and environmental parameters, facilitating the development of more efficient and effective OA organoid construction strategies. Furthermore, AI-powered optimisation significantly reduces variability in traditional cell culture processes, thereby enhancing reproducibility and consistency in organoid production. Natsume et al. developed a robotic AI system incorporating a batch Bayesian optimisation algorithm to refine differentiation protocols for iPSCs-derived retinal pigment epithelial cells, systematically optimising cell culture conditions to achieve superior differentiation outcomes [239]. Similarly, Su et al. utilised machine learning to analyse the spatial structure of hydrogels, predicting cellular behaviour dynamics and assessing the impact of external stimuli (e.g., temperature, chemical agents and enzymatic degradation) on hydrogel performance. This AI-driven approach enables the identification of optimal hydrogel synthesis strategies for biological applications [240]. In the context of OA organoid manufacturing, AI can be effectively integrated with pathological OA organoid biocharacteristics to systematically optimise: Culture media composition and gradient control; Mechanical stimulation loading patterns; Passaging cycle parameters. By establishing standardised operating procedures, AI ensures precision and reliability in OA organoid culture. Additionally, real-time monitoring technology is pivotal in culture condition optimisation. By integrating sensors into the culture system, it is possible to dynamically monitor key microenvironmental parameters, such as oxygen concentration, pH levels and nutrient availability. These real-time data are processed by AI algorithms, which autonomously adjust culture conditions to maintain optimal cellular growth states. This closed-loop feedback mechanism enhances experimental accuracy and minimises human intervention errors, thereby improving the overall efficiency and reproducibility of OA organoid manufacturing [241].

5.1.3 Image Processing System

In recent years, deep learning technology has been widely applied in the diagnosis and research of OA. To support intelligent manufacturing and data-driven analysis, it is essential to establish a high-quality multi-modal organoid image database. Deep learning-driven image denoising and 3D reconstruction algorithms can effectively integrate data from multiple imaging modalities, such as X-ray, CT scans and MRI, forming a comprehensive database containing over 100,000 annotated images. Standardised preprocessing pipelines are implemented to mitigate image blurring and registration artefacts caused by light scattering, ensuring high-quality datasets for subsequent quantitative analyses [242, 243]. Leveraging this large-scale multi-modal image database, an improved U-Net++-organoid segmentation algorithm has been developed to precisely quantify morphological parameters, including volume, surface area and sphericity. Furthermore, by integrating Haralick texture features and vascular network topology analysis, critical pathological phenotypes such as synovial hyperplasia and necrotic region distribution in OA can be extracted. Philippe et al. conducted a longitudinal study using 3D MRI data of OA patients, applying deep learning models for MRI analysis to predict disease progression and prognosis [244]. These extracted features not only aid in understanding organoid developmental abnormalities but also provide key insights for early diagnosis and treatment strategies [245, 246]. For implementation, Savani et al. utilised pretrained models, including ResNetV2 and Xception, for data classification and feature recognition. A visualisation platform was developed to integrate growth trajectory tracking and phenotype evolution heatmaps [243]. By combining morphological parameters with functional biomarkers, this approach provides intelligent decision support for high-throughput experiments.

5.1.4 Multi-Omics Deep Neural Network System

The development of a multi-omics deep neural network system aims to analyse the complex relationships between multi-modal organoid data (genomics, metabolomics, imaging) and culture condition parameters (e.g., growth factors, mechanical stimulation, oxygen levels). By integrating genomics, metabolomics and imaging datasets, this system captures dynamic molecular changes and phenotypic features during organoid development. Genomics data reveal key gene expression regulatory mechanisms. Metabolomics data provide temporal profiles of metabolite concentrations. Imaging data reflect morphological and structural changes in OA organoids [247]. Built on a deep learning framework, this system employs multi-modal variational encoders to unify diverse datasets into a shared latent representation, with decoders generating predictive outputs. By integrating single-cell transcriptomic dynamics, metabolite concentration shifts and microscopic imaging features, it establishes a quantitative correlation model between organoid developmental states and culture conditions [248]. This model deciphers the impact of key culture parameters (e.g., growth factor concentration, mechanical stress, oxygen gradients) on organoid differentiation and functional maturation, forming a data-model-experiment closed-loop system. Huang et al. proposed Multi-Omics Graph Convolutional NETwork, which integrates mRNA, DNA methylation and miRNA data to classify renal carcinoma with higher accuracy than traditional methods [249]. This framework can be extended to OA organoids, optimising graph convolutional network layers and view correlation discovery networks to analyse multi-omics responses to culture conditions. To address the small-sample data limitation, this system incorporates a reinforcement learning-based culture condition optimisation algorithm. RL mimics human decision-making processes, continuously refining strategies through trial and error. In OA organoid culture, RL algorithms can simulate dynamic experimental conditions, predicting optimal culture parameter combinations. Additionally, by integrating generative adversarial networks (GANs), high-quality synthetic datasets can be generated to compensate for real-world data scarcity. GANs train generators and discriminators, producing synthetic data distributions that closely resemble real biological datasets, thereby enhancing model generalisability [250, 251]. Carlo et al. proposed a bio-manufacturing optimisation method combining reinforcement learning and simulation, in which environmental state observations, action generation and reward collection iteratively refine optimal biomanufacturing strategies, ultimately generating optimised target sequences [252].

5.2.1 Microfluidic Equipment

In the ‘human-free factory’ intelligent manufacture, the microfluidic microporous array chip can be constructed in a modular manner, allowing for precise and dynamic design of the stem cell culture environment. This enables comprehensive simulation of the cellular components of the inflammatory microenvironment, inflammatory factor composition and immune factors, ultimately creating organoids that accurately reflect the pathological characteristics of OA. Its core functions and advantages include: Cell Capture and Aggregation: The precise microwell array design enables effective cell capture and aggregation, establishing a stable physical foundation for organoid formation. For instance, hexagonally arranged microchambers (≥ 1000 wells per chip) optimise cell distribution. Fibronectin-coated microwell surfaces enhance cell adhesion, thereby improving culture efficiency [253]; Organoid Size Control and Standardisation: Microfluidic chips simulate fluid shear forces and optimise microwell spacing and bottom micro-pillar structures, ensuring high uniformity in organoid size. This design minimises batch-to-batch variability, providing a reliable framework for standardised OA organoid manufacturing [254]; Integration of Miniature Biosensor Arrays: The microfluidic chip is embedded with miniaturised biosensor arrays, which, when combined with the real-time monitoring and feedback system, enable continuous tracking of key culture parameters (e.g., oxygen concentration, temperature). These sensors dynamically regulate culture conditions, ensuring precise control over organoid development. This real-time sensor integration enhances mechanistic studies of OA and provides valuable feedback for disease modelling [255]; Development of Modular Culture Platforms: The automated organoid culture platform, based on microwell array technology, supports high-throughput, parallelised and standardised organoid production. Each module contains six independent microfluidic chips, equipped with miniature heating units and gas exchange membranes, enabling simultaneous multi-batch culture. This modular approach improves experimental efficiency while significantly reducing production costs [254].

5.2.2 Automated Liquid Handling Equipment

With the increasing demand for biomedical research and clinical applications, automated liquid handling equipment is pivotal in the scalable production and precise management of organoids. These equipment integrate multiple functions, including cell seeding, passaging, cryopreservation, media exchange and real-time monitoring, enabling fully automated workflows from cell culture to organoid generation. By evaluating batch-to-batch variability coefficients, these equipment validate product stability and enhance the efficiency of large-scale organoid manufacturing. Schimitt et al. proposed an automated organoid culture method known as AutoCRAT, which enables large-scale automated production of MSCs, chondrocytes and extracellular vesicles derived from induced iPSCs [256]. The core components of the automated liquid handling equipment include: High-Precision Robotic Arms: The equipment employs high-precision robotic arms to ensure accurate and reproducible cell seeding and handling. These robotic arms precisely execute cell positioning and manipulations, minimising human error and improving experimental reproducibility. Ioannis et al. developed a robotics-based automation strategy for media exchange and imaging of cartilaginous microtissues cultured in static microwell platforms, demonstrating the efficiency and consistency of robotic-assisted handling [257]; High-Accuracy Syringe Pumps: The equipment is equipped with syringe pumps with flow rate errors below 1%, enabling precise control over liquid dispensing. This precision is particularly critical for media replacement and cryopreservation solutions, ensuring stable and reproducible culture conditions [258]; Miniaturised Cryopreservation Module: The cryopreservation module features a rapid cooling rate of 10°C/min, designed for gradient cooling with dimethyl sulfoxide (DMSO) during cell cryopreservation. This function enhances cryopreservation efficiency while minimising cellular damage during the freezing process [256].

5.2.3 Multi-Parameter Sensing Equipment

In the ‘human-free factory’ intelligent manufacturing strategy, the multi-parameter sensing equipment is a core technology enabling real-time monitoring and intelligent regulation of the organoid microenvironment. This equipment integrates miniaturised sensors for pH, oxygen partial pressure, metabolites (e.g., lactate/glucose) and mechanical stress detection, constructing a multi-dimensional dynamic data acquisition platform for real-time environmental monitoring. These sensors operate at millisecond-level temporal resolution, continuously tracking key biophysical and biochemical parameters of organoids. Key Components of the Multi-Parameter Sensing Equipment include: pH and Oxygen Partial Pressure Sensors: The detection of pH and oxygen levels primarily relies on fluorescent fibre-optic sensor technology. Sensor probes are coated with Ru(dpp)₃ ⁺ oxygen-sensitive films, which emit fluorescence signals in response to oxygen concentration fluctuations, enabling real-time oxygen monitoring [259]. Additionally, pH-sensitive visualisation sensors based on metal–organic frameworks and covalent organic frameworks have been extensively studied. These sensors are designed to enhance pH measurement accuracy and emit multi-wavelength fluorescence signals, improving detection sensitivity [260]; Metabolite Monitoring: The monitoring of metabolites (e.g., lactate and glucose) is achieved using microfluidic electrochemical chips. These chips incorporate glucose oxidase and lactate dehydrogenase-modified electrodes, allowing for precise metabolite concentration detection. This enables researchers to assess the metabolic state of organoids under specific culture conditions [261]; Mechanical Stress Sensors: The equipment integrates a flexible piezoresistive sensor array, designed to detect compression stress within the organoid microenvironment. These sensors feature a measurement range of 0–20 kPa and a response time of less than 10 milliseconds, enabling real-time tracking of mechanical forces acting on the organoid. This is critical for simulating the mechanobiological stimuli present in the pathophysiological environment of OA [262]. Building upon the multi-parameter sensing equipment, researchers have further developed an intelligent regulation equipment. This equipment utilises a microfluidic valve array to dynamically adjust: Culture medium perfusion rates; Growth factor release gradients; Mechanical stress loading patterns. By comparing static and dynamic culture equipments, studies have demonstrated that dynamic regulation significantly enhances organoid maturation and functional properties. This equipment serves as a technological foundation for simulating in vivo pathological microenvironments, advancing next-generation organoid-based disease modelling [259].

5.2.4 Imaging Equipment

In recent years, with the rapid advancement of biomedical imaging technologies, multi-photon microscopy (MPM) and light-sheet fluorescence microscopy (LSFM) have demonstrated tremendous potential in organoid research due to their unique imaging advantages. To overcome the depth and resolution limitations of traditional imaging techniques, a combined multi-photon confocal microscopy and light-sheet imaging equipment has been developed, enabling cross-scale simultaneous capture and measurement of organoid 3D structures and subcellular dynamics. This equipment integrates the strengths of both MPM and LSFM and allows rapid switching between imaging modalities via a customised sample stage. Multi-photon confocal microscopy utilises laser excitation wavelengths of 690–1300 nm, achieving a Z-axis resolution of up to 10 μm. Light-sheet fluorescence microscopy employs dual-laser excitation at 488/561 nm, providing an XY resolution as high as 0.5 μm. This design significantly enhances imaging depth, reduces phototoxicity and enables high-resolution imaging of deep tissues [263]. Additionally, by incorporating a piezoelectric ceramic-driven 3D sample translation equipment, this setup supports long-term continuous observation and automated multi-field stitching, ensuring reliable imaging for over 24 h in studies involving cells, organoids and live tissues. Steventon et al. successfully applied this continuous observation and automatic stitching equipment in long-term zebrafish studies [264]. To further improve deep-tissue imaging quality, an adaptive optics module has been introduced into the optical path of the light-sheet microscope. This module compensates for aberrations caused by light scattering within organoid structures, ensuring clear and accurate deep imaging. Such optimisation not only enhances imaging quality but also provides critical support for analysing complex biological structures. Luke et al. demonstrated this by employing an adaptive optics module for imaging mitochondria through bone in live mice using two-photon fluorescence microscopy [265].